In a push-to-talk communication environment, a plurality of users or speakers joins a common channel, for example a VTG (Virtual Talk Group) to communicate with one another. Typically, the communication channel is configured such that only one speaker is allowed to speak at a time. Thus, speech which is audible in such a channel generally comprises a plurality of media segments (e.g. portions of speech) from respective speakers which media segments are appended serially one media segment after another. The communication in such a push-to-talk environment is therefore generally ordered and is suitable for safety and security operations.
Speech of safety and security operations is usually recoded in order to facilitate forensic analysis of events. The same recording can be used by latecomers who join the operation or session (e.g. log onto the VTG) after it has started, in order to inform or notify the latecomers about what has previously transpired. Operations are usually managed by one or more “principals”. This individual is generally the highest ranking person present, or a specialist who is recognized for his understanding or authority; usually what he says carries the key actions or content. As a new user joins an operation, he or she typically wants to understand what had previously transpired in the event.
The user can invoke the replay mechanism and listen to the replay of all that had been said prior to his joining. If the new user is pressed for time, he may choose to listen only to the media segments (e.g. voice clips or speech portions) of the principals. This, however, has the disadvantage that he could miss a comment or question from one of the other speakers. The user may speed up the whole replay, but this may detract from his ability to focus on the principal's messages. Yet another option is to modify the replay speed continually, for instance slowing down the voice of the principal and speeding up the reply of the spoken statements of the other speakers. This may shorten the time required to listen to the recorded message but may not be practical when the new user needs to cater to unfolding events.